It is of great practical significance for the state of charge (State of Charge, SOC). At present, the application of a single SOC estimation method can not get the optimal SOC estimate value well, resulting in a large deviation of the estimation. In view of the above problems, the joint application scheme of AH + EKF is proposed. First, the first-order RC circuit model was selected for parameter identification based on the battery discharge data, and then using the Beijing electric bus measured data (Bulk Battery Data System Test, BBDST) compared several estimation method, when a single AH integral method and EKF algorithm, the maximum error of 17.7% and 20.2%, respectively, using the method of AH and EKF algorithm overall maximum error is not more than 1%, the joint application of the algorithm has a very good estimation effect.